CNN genomic predictor (Conv1D + GAP + dense, base R)
Usage
morie_cnn_genomic(
x,
y,
markers,
n_filters = 8,
kernel = 3,
hidden = 8,
n_epochs = 150,
lr = 0.01,
l2 = 0.001,
seed = 0,
deterministic_seed = NULL
)Arguments
- x
Optional fixed-effect design.
- y
Numeric response.
- markers
(n x m) genotype matrix.
Hyperparameters.
- deterministic_seed
Optional integer; if supplied, RNG state is derived via
morie_det_rng()keyed on ("cnnge", deterministic_seed) so Py<->R streams agree on the canonical fixture. WhenNULL(default) behaviour is unchanged.
Examples
morie_cnn_genomic(x = rnorm(50), y = rnorm(50), markers = matrix(sample(0:2, 200, TRUE), 50, 4))
#> $estimate
#> [1] 0.04161699
#>
#> $y_hat
#> [1] -0.266923988 0.012658896 0.032333050 -0.002983763 0.237756522
#> [6] -0.054232818 0.094833185 0.094833185 0.014041034 0.050907718
#> [11] -0.008534918 -0.055510944 0.072393306 -0.055510944 0.094384232
#> [16] 0.139912742 0.024328512 0.149864483 0.237756522 -0.118523966
#> [21] -0.002983763 -0.085088989 0.139912742 -0.013745582 -0.119486518
#> [26] 0.237756522 -0.085088989 0.080486116 -0.052919952 -0.053155050
#> [31] -0.016757522 -0.137624885 0.237756522 -0.247147181 -0.022397529
#> [36] 0.298763318 0.094833185 0.353186472 0.298763318 0.080486116
#> [41] 0.260280328 0.023278348 0.177940848 0.026214398 -0.054232818
#> [46] -0.055510944 0.032333050 -0.023203537 -0.054705824 0.069125486
#>
#> $W_conv
#> [,1] [,2] [,3] [,4] [,5] [,6]
#> [1,] -0.08149808 -0.2433687 0.2237267 -0.6955361 -0.5620651 0.2256334
#> [2,] 0.86612501 -0.5355294 -0.4575499 -0.5987644 0.2331776 0.9545705
#> [3,] -0.41785978 -0.1063195 0.4556717 0.8392138 -0.2494119 0.8994827
#> [,7] [,8]
#> [1,] -0.1903252 0.39614611
#> [2,] -1.3180241 0.34586332
#> [3,] 1.4330395 0.02252572
#>
#> $b_conv
#> [1] 0.0050549012 0.0063367307 -0.0159569727 0.0005670042 -0.0004568030
#> [6] -0.0189098326 0.0006100363 0.0228631854
#>
#> $W1
#> [,1] [,2] [,3] [,4] [,5] [,6]
#> [1,] 0.19471654 -0.442113376 0.64761323 0.182995138 0.37402693 0.34523697
#> [2,] -0.06617702 0.226611525 -0.12078290 -0.009595931 0.07475823 0.05213002
#> [3,] 0.13185761 -0.015810014 -0.56951162 -0.020967444 -0.31640105 0.15165338
#> [4,] -0.12251293 -0.611640194 0.08039668 0.201164085 0.41251910 -0.04261144
#> [5,] -0.45178079 0.001095636 0.10444716 -0.453371992 -0.69573162 -0.10658647
#> [6,] 0.45356432 -0.221314514 -0.30674127 0.304314853 -0.14045253 0.14945745
#> [7,] 0.51642698 -0.120394098 -1.01822034 0.028475353 -0.10158035 0.41013739
#> [8,] -0.08285673 -0.407966471 -0.21314254 0.231601603 -0.04613723 -0.87014655
#> [,7] [,8]
#> [1,] 0.21616712 0.10026578
#> [2,] 0.12569088 -0.15169197
#> [3,] -0.15218808 -0.42211699
#> [4,] 0.34455235 -0.11614043
#> [5,] -0.12416736 -0.32671325
#> [6,] -0.03858209 -0.06726747
#> [7,] 0.30290346 0.13471698
#> [8,] 0.62269542 -0.29497731
#>
#> $b1
#> [1] -0.0249024261 0.0001048920 0.0022197498 0.0310179228 -0.0001277762
#> [6] -0.0050783158 0.0021893505 -0.0004068752
#>
#> $w2
#> [1] 0.96983010 -0.01892357 0.37652106 -0.75067367 0.21787081 -0.46978956
#> [7] 0.34591197 0.10165182
#>
#> $b2
#> [1] 0.03422313
#>
#> $loss_curve
#> [1] 0.9283070 0.9233297 0.9187637 0.9145667 0.9107005 0.9071313 0.9038286
#> [8] 0.9007657 0.8979184 0.8952654 0.8927873 0.8904669 0.8882888 0.8862393
#> [15] 0.8843061 0.8824781 0.8807454 0.8790991 0.8775314 0.8760352 0.8746039
#> [22] 0.8732320 0.8719143 0.8706461 0.8694233 0.8682422 0.8670995 0.8659920
#> [29] 0.8649171 0.8638724 0.8628556 0.8618647 0.8608979 0.8599537 0.8590304
#> [36] 0.8581269 0.8572419 0.8563743 0.8555231 0.8546874 0.8538664 0.8530594
#> [43] 0.8522656 0.8514844 0.8507153 0.8499578 0.8492113 0.8484755 0.8477498
#> [50] 0.8470341 0.8463278 0.8456307 0.8449425 0.8442629 0.8435917 0.8429286
#> [57] 0.8422734 0.8416259 0.8409859 0.8403532 0.8397277 0.8391091 0.8384973
#> [64] 0.8378922 0.8372936 0.8367015 0.8361156 0.8355358 0.8349621 0.8343943
#> [71] 0.8338324 0.8332761 0.8327255 0.8321804 0.8316407 0.8311064 0.8305773
#> [78] 0.8300534 0.8295346 0.8290208 0.8285120 0.8280080 0.8275088 0.8270143
#> [85] 0.8265245 0.8260393 0.8255585 0.8250823 0.8246104 0.8241429 0.8236797
#> [92] 0.8232206 0.8227658 0.8223150 0.8218683 0.8214256 0.8209868 0.8205519
#> [99] 0.8201208 0.8196935 0.8192700 0.8188502 0.8184340 0.8180214 0.8176123
#> [106] 0.8172068 0.8168047 0.8164060 0.8160107 0.8156187 0.8152300 0.8148446
#> [113] 0.8144624 0.8140833 0.8137074 0.8133345 0.8129648 0.8125980 0.8122342
#> [120] 0.8118734 0.8115154 0.8111604 0.8108081 0.8104587 0.8101121 0.8097681
#> [127] 0.8094269 0.8090884 0.8087525 0.8084192 0.8080885 0.8077626 0.8074431
#> [134] 0.8071261 0.8068114 0.8064990 0.8061890 0.8058813 0.8055758 0.8052726
#> [141] 0.8049716 0.8046727 0.8043761 0.8040816 0.8037891 0.8034988 0.8032106
#> [148] 0.8029244 0.8026402 0.8023580
#>
#> $se
#> [1] 0.895588
#>
#> $n
#> [1] 50
#>
#> $method
#> [1] "Conv1D + GAP + dense (base R)"
#>